IR Remote and STM32 using ChibiOS

IR remote

Using Infra-Red for remote control is a technology born in the 70’s and after 40 years it is still used because the good ratio performances-costs and low power consumption. IR remote require a line of sight, because of that latest remote includes bluetooth technology.

Today it is possible to buy a small remote and receiving circuitry spending a few dollars. In Embedded system using IR could be a fast and cheap solution for remote control even if nowadays RF, Bluetooth and WiFi modules are very affordable.

Lack of documentation

Searching documentation for cheap devices is often not easy at all. We cannot provide a reference manual nor datasheet for IR remote we are going to use (and even if we could most likely reader is trying to use a different device). So we want to collect here every information gathered on the internet trying to put them together pursuing our purpose: using an IR Remote.

Communication occurs through Infra Red pulses. Information is encapsulated in pulse widths: this technique is also know as PWM. We often talked about PWM on this website even if we almost used it as voltage regulation technique. Indeed we already used it ad digital communication modulation method getting data from DHT11.

Modulated signal
An example of a digital signal modulated over a square wave carrier.

The PWM is then moved from baseband to an higher frequency depending on device. This procedure goes under the names of modulation and results in beneficial effects like noise rejection, reduction of losses during signal propagation, ability of propagate multiple signal on the same channel (air in this case). The result of this new modulation is a non-baseband PWM (See Fig.1)

Typically, carrier frequency (i.e. frequency in which is centred PWM signal after modulation) are in range of 36kHz~40kHz. If we use an IR diode (and related gain amplifier) as receiver reading signal with an oscilloscope we will see a signal similar to red signal of Fig.1. We can find some very cheap IR receiver that embeds the IR diode, a gain amplifier, a tuned filter and a Schmitt trigger. Since this receiver has a tuned filter we can find different devices for different carriers (36kHz, 36.5kHz, 37kHz, etc.).

TSOP1838 application circuit
A typical application circuit from TSOP1838 datasheet.

Most IR remote kit provides an IR receiver with a small PCB in which few needed passives are housed. These passives typically are a pull-up resistor and an electrolytic capacitor to filter noise on power supply, they are required specially if cables are long (see Fig.2). For the remote of this article frequency carrier is 38 kHz. Note that a IR receiver typically has three pins:

  1. VCC, connection to power supply (typically 2.5~5.5V, but since output depends on it we suggest 3.0V);
  2. GND, connection to ground;
  3. OUT, baseband PWM signal.

Signal on the output terminal is the envelop of IR signal i.e the baseband PWM signal (Referring to Fig.1 it is similar to the black one).

Reverse engineering

As said, information is encapsulated in square wave width. Now we need to decode signals. Proposed method is reverse engineering: we can mount circuity and write some minimal code printing on a terminal captured widths. As for DHT11 we can use ICU driver for this task, main will be something like this:

Using that code we can understand certain mechanism behind the communication. As example here a copy of some test made with my IR remote

Note that we can recognise certain repetitions. Here I pushed 8 buttons keeping pressed the last one. We can easily recognise that:

  1. every command is identified by a start pulse (45/46 counts i.e. 4,5 ms since ICU clock frequency is 10kHz) and an end pulse (39.4/39.5 ms);
  2. commands are composed by 32 bit where shortest (0.6 ms) is ‘0’ and longest (1.7/1.8 ms) is 1;
  3. communications are divided by a comma pulse (2.3 ms);
  4. commands are dead pulses between two communication (10769, 8680, 10293 counts, etc.);
  5. when button is kept pressed we receive a pulse (95.9/96 ms) followed by a comma that simply means repeat last command.

So we can ignore dead pulses, use start and stop to identify a command and properly recognise repeat command. Last step before write a demo is decode every button.

Proposed demo

Demo explained

The core of our demo is the ICU width callback. In this callback we have to compose the command value or detect the repeat flag. To understand following code let’s list some certitudes:

  • both command and flag are completed on comma pulse;
  • when we detect a start pulse, we must be ready to compose command value in the next 32 pulses;
  • when we detect an end pulse the command should be composed;
  • long dead pulse are useless to our purpose;

The callback above decodes IR signal and broadcast a flag on comma. Note that there are some simple check mechanism to detect uncompleted or wrong communication: as example command value is composed on a temporary variable and is copied on command only on end pulse if start pulse as been detected and 32 pulses occurred.

Concluding main() wait for events printing last command on a base sequential stream:

Project download

The attached demo has been tested under ChibiOS 17.6.x. Note that you can find more recent version of this project int the Download Page. Note also that starting from the version 20 all the demos from PLAY Embedded will be distributed together with ChibiStudio.


Be the first to reply at IR Remote and STM32 using ChibiOS

Leave a Reply